Fujitsu Laboratories' new architecture rivals quantum computing

The architecture is capable of handling more than current quantum computing machines, Fujitsu says.

Fujitsu  Laboratories announced this morning that it has teamed up with the University of Toronto to build a computer architecture that's basically faster than quantum computing in utility. According to the company's press release, the machine is designed to solve 'combinatorial optimization problems'. That basically means it is capable of finding the best possible combination of elements, out of a countless number of elements.  

Through the employment of conventional semiconductor technology, with flexible circuit configurations, the machine is capable of handling more than current quantum computing machines. Fujitsu said it implemented a prototype using FPGAs (field-programmable gate array) and discovered that the architecture can perform computations 10,000 times faster than a conventional machine. Fujitsu hopes this new architecture will help speed up streamline, improve post-disaster recovery plans and be able to help formulate economic policies. 

“It will also make possible the development of new ICT services that support swift and optimal decision-making in such areas as social policy and business, which involve complex intertwined elements,” it said in a press release. 

Expanding the bit scale of the technology, the architecture can be used to quickly solve tough combinatorial optimisation problems like optimising distribution to thousands of locations, or optimising profits from multiple projects while on a tight budget.  By fiscal 2018, the company wants to have a prototype capable of handling real-world problems of 100,000 bits to one million bits.

Image Credit: Ricochet64 / Shutterstock